Salinity, Water Level, and Flow Considerations for Assessing Environmental Water Requirements of the Lower Latrobe River

7th Australian Stream Management Conference - Full Paper

1 2 3 Lauchlan Arrowsmith C , Brizga S and Keogh E

1. Water Technology, 15 Business Park Drive, Melbourne, 3133. Email: [email protected] 2. S. Brizga & Associates Pty Ltd, Melbourne, Victoria, Australia. 3. West Gippsland Catchment Management Authority, Traralgon, Victoria, Australia.

Key Points • Estuary Environmental Flows • Salinity impacts on environmental flow recommendations • Freshwater flows to manage salinity as well as water levels and inundation • Detailed 3D hydrodynamic modeling to set salinity flushing criteria

Abstract The lower Latrobe River is an estuarine system which connects the catchments of the Thomson, Macalister and Latrobe Rivers with Lake Wellington and the Gippsland Lakes. The estuarine conditions of the river are unusual, as its distance from the ocean entrance of the Gippsland Lakes results in a micro-tidal water level range, with water levels in the Lake Wellington and lower Latrobe River dominated by winds, waves and atmospheric pressure as well as the effects of upstream river inflows. Salinity conditions are also highly variable, ranging from predominantly fresh to almost ocean salinity during the recent extended drought period. Salinity is a crucial factor with regard to the ecological condition of the system and an environmental flows study investigated how environmental water could be used to moderate the effects of increasing salinity regimes as well as meet the ecological objectives of the system as a whole. Hydrodynamic modelling was used to assess the flushing behaviour of freshwater inflows on the lower river and this behaviour was then linked back to the ecological requirements. Environmental flow recommendations were developed which focused on freshwater inflows to manage salinity, along with more typical criteria around flow inundation in terms of water levels and frequency. Quantification of these recommendations was made possible through the detailed hydrodynamic modelling.

Keywords Estuary, Salinity, Environmental Flows, Hydrodynamic Modelling

Introduction The Latrobe estuary is the outlet for Thomson, Latrobe and Macalister catchments. It drains a total catchment area of 8,600 km2, including the 5000 km2 Latrobe River catchment and 3,600 km2 Thomson River catchment. The major rivers and many key tributaries rise in the Great Dividing Range. Southern tributaries of the Latrobe River, such as Traralgon Creek, rise in the Strzelecki Ranges. After emerging from the mountains and ranges, the rivers then flow across the Gippsland plains before discharging into Lake Wellington.

The Latrobe estuary is a component of a much larger estuarine complex – i.e. the Gippsland Lakes. The Latrobe River flows into Lake Wellington, a large estuarine lagoon. The Gippsland Lakes drain to the ocean (Bass Strait) via a permanent artificial entrance at Lakes Entrance that was constructed in the late nineteenth century. The former natural entrance was variable, subject to periodic opening and closing as well as shifts in location.

The Latrobe estuary study area has a temperate climate with a late winter-early spring dominant rainfall regime. It experiences cold humid winters and warm dry summers. Snowfalls regularly occur on the higher peaks in Great Dividing Range in winter and snowmelt contributes to seasonal streamflow patterns. The seasonal streamflow regime reflects these climatic influences, with the highest streamflows occurring in late winter / early spring, and the lowest streamflows occurring in late summer / early autumn.

The tides at the ocean entrance of the Gippsland Lakes are semidiurnal and microtidal, with a spring tide range of around ±0.7m. However, the tides at the mouth of the Latrobe estuary are significantly damped due to the distance of

Lauchlan Arrowsmith,C., Brizga, S., and Keogh.E. (2014). Salinity, Water Level, and Flow Considerations for Assessing Environmental Water Requirements of the Lower Latrobe Valley, in Vietz, G; Rutherfurd, I.D, and Hughes, R. (editors), Proceedings of the 7th Australian Stream Management Conference. Townsville, Queensland, Pages 58-65. 58 7ASM Full Paper

Lauchlan Arrowsmith et.al. – Salinity. Water Level, and Flow Considerations of the Lower Latrobe River the river from the Lakes entrance, with astronomical tides leading to around 0.03 m of variations in water levels at the river mouth. More important influences on the water level regime of Lake Wellington, which controls water levels in the Latrobe estuary, include barometric pressure and seiching due to the effects of wind on the lake surface. Although the main tidal constituents are minor the level in Lake Wellington is quite variable; the average level is around 0.16 m, with a 95% confidence interval range of -0.11 to 0.44m.

The Latrobe Estuary is experiencing significant pressures on its condition arising from the impacts of human activity (Earth Tech, 2005). On the broad catchment scale, extensive water resource development has significantly reduced the flows in the main inflowing rivers and direct modifications to river and creek channels such as channel straightening and desnagging has occurred. For example, the main stem of the Latrobe River below Lake Narracan was shortened by about one-fifth as a result of approximately 72 artificial meander cut-offs, some of which have more recently been reinstated (Earth Tech 2005, Alluvium 2009). Major de-snagging was also carried out (Earth Tech 2005). These catchment scale changes and activities affect the quantity and quality of riverine inflows to the Latrobe estuary, as well as providing sources of exotic flora and fauna.

Locally, cattle grazing is associated with stock ingress to the river from unfenced frontages and loss of phragmites due to grazing. Other local pressures on the condition of the Latrobe estuary arise from channel modifications (desnagging and dredging) and boat traffic. Localised bank stabilisation works have been undertaken by West Gippsland Catchment Management Authority (WGCMA) to protect key assets, such as remnant red gums, from bank erosion, particularly along the Thomson River between the Sale Canal and Latrobe River confluence.

And crucially, under natural conditions, the Gippsland Lakes system was subject to intermittent mouth closure and opening, but with the construction of the permanent mouth at Lakes Entrance in 1889, it is always subject to sea water influence. This has altered the hydrodynamics of the Gippsland Lakes, including increased salinity. Lake Wellington has changed from a freshwater – brackish water body, to brackish – saline (e.g. Longmore, 1990; Gippsland Lakes Tasks Force, 2004; Ecos Consortium, 2009). This has had profound ecological implications.

Study Approach

Overview The physical character of the Latrobe estuary is determined by its geomorphology, hydrology and hydrodynamics. The physical characterization of the system in terms of hydrologic and hydrodynamics involved the development of detailed 1D and 3D hydrodynamic models of the estuary to quantify the key relationships between flow, water level and salinity throughout the system. Geomorphological and ecological characterization involved the developed of conceptual models of the estuarine ecosystems based on the physical characterization. Environmental assets were then identified, and environmental objectives were developed for these assets based on relevant legislation, policy and existing plans. This was followed by the identification of environmental water requirements for the environmental assets and objectives across the estuary. For further details refer Brizga et al (2013).

Environmental flow recommendations were developed to address the environmental objectives, based on the detailed understanding of the system gained during the study, including an assessment of the sensitivity and environmental risk associated with these requirements across a range of hydrological and water related conditions including possible future climate and water use scenarios. The associated shortfall and compliance of these recommendations for the range of hydrological and water related conditions were then assessed.

This paper focusses on the detailed physical characterization of the estuary, in particular how the hydrodynamic modeling was used to understand the implications of salinity on ecological character and the how the flushing behaviour of freshwater inflows could be supplemented by the use of environmental water.

Hydrodynamic Modelling Scope The purpose of the 1-D hydrodynamic model of the estuary was to quantify the bankfull capacity of the estuary channel and to characterise the flow/level relationship throughout the estuary to support the interpretation of ecological processes. Because of the complex relationship between flows and levels in the Latrobe estuary and the adjacent

7ASM Full Paper

Lauchlan Arrowsmith et.al. – Salinity. Water Level, and Flow Considerations of the Lower Latrobe River wetlands the 1-D model was used to provide the following key information based on synthetic natural and current daily time series of river flows developed for the study:

• An understanding of the flow and level relationships throughout the study area; • Information on how the variation in water level in Lake Wellington affects these relationships; • Characterisation of the flow and water level in terms of frequency, duration, timing; • Characterisation of salinity in the estuary (flushing events) developed from the 3-D modelling.

For this analysis a synchronous time series of river inflows and measured Lake Wellington water levels was used as the level in Lake Wellington is dominated by atmospheric pressure and wind effects rather than tide.

The 1-D model was then also used in later stages of the study to assess the relative impacts of each water resource, environmental entitlement and climate change scenarios on the basic flow statistics developed for natural and current flow conditions, as well as flow thresholds relating to salinity developed through the 3-D modelling.

The purpose of the 3-D hydrodynamic model of the estuary was to understand the estuary tidal dynamics and salinity structure, and its sensitivity to variations in inflow discharges. Once calibrated the following scenarios were tested:

• Low inflow conditions (to assess the impact of saline intrusion from Lake Wellington and the movement of the salt wedge through the estuary); • Flushing flows (to assess the river inflows required to flush saline water out of the estuary). The outputs for each simulation included time series variation of vertical salinity profiles, variation in velocity with depth throughout the estuary and characterization of the salt wedge in terms of location and likely depth to the halocline.

Physical Characterization

Water Level and Flow The hydraulic relationship between river inflow and water level (Q-H) is non-linear throughout the study area. This can clearly be seen in the Figure 1, which shows the range in water levels that can occur just downstream of the confluence of the Thomson and Latrobe Rivers (known as the Swing Bridge) for any given combination of river inflows. The scatter in the Q-H relationship was found to be greater at the downstream end of the system, where water levels are strongly controlled by the level of Lake Wellington and the backwater effects of flows into the fringing wetlands are more dominant.

Figure 1. Water Level versus Flow Relationship at the Swing Bridge

7ASM Full Paper

Lauchlan Arrowsmith et.al. – Salinity. Water Level, and Flow Considerations of the Lower Latrobe River

Flow exceedance curves were developed for eleven locations in the estuary and were interpreted to understand changes to the flow regime between natural and current conditions. The natural and current conditions were defined as follows (SKM, 2011):

• Natural - the natural sequence of flows that would have occurred over the historical climate conditions over the period (1957-2007) without extractions and discharges of water in the catchment. The effects of other land use changes such as urbanisation are not considered.

• Current - the sequence of flows that would have occurred over the historical climate conditions over the period (1957-2007) based on a defined level of water development demand and infrastructure.

The effect of the changes in water resource development between the two conditions is greatest on the Thomson River, where all flow frequencies were reduced by up to 74% (maximum change) between natural and current scenarios. For the Latrobe River the maximum reduction was 27%. At the Swing Bridge, downstream of the confluence between the Thomson and Latrobe Rivers, flow frequencies were reduced by up to 43%.

As the relationship between flow and water level is not easily defined throughout the study area the effect of changes to inflows on water level are also highly variable. Water levels in the Thomson River are more affected by development in the catchment than the Latrobe River, as would be expected given the greater hydrologic impacts of water resource development plus the narrow deep channel morphology. At the Upper Latrobe location the effect is less pronounced, predominantly because there is less development in the Latrobe River catchment. Water resource development has little effect on water level regime in the lower reaches due to the dominant effect of Lake Wellington on water levels.

Salinity The salinity of the Lower Latrobe estuary is directly related to the salinity conditions in Lake Wellington. Historically the lake was predominantly fresh, however with the permanent entrance to the Gippsland Lakes the salinity in the lake and the estuary has increased.

Saline waters from Lake Wellington move up into the Latrobe estuary depending on the river flow conditions. Just as Lake Wellington can become flushed of saline water when flows are above the threshold of around 130 GL/month, the ability of the salt wedge to penetrate throughout the estuary depends on the incoming river flows.

To understand this link between river inflows and salinity condition in the estuary a range of flow conditions were simulated in the 3D model. From these results the likely location of saline water for a given flow were plotted, allowing flushing flow conditions at the various reporting points throughout the estuary to be developed. The model also allowed insight into the vertical position of the halocline in the water column.

To facilitate data interpretation and linkages to the ecological characterisation, the water column was conceptualised as two layers that provide a basis for the quantification of hydraulic and salinity properties of the water column (Figure 2). This simplified representation enables discrimination of stratified conditions (upper freshwater layer and lower more saline layer) from conditions where the salt wedge is absent (both layers fresh) or where the water column is mixed (both layers brackish or saline).

The cross-sectional profile used in Figure 2 represents the typical morphology of the channel cross-section of the Thomson River downstream of the Sale Canal and Latrobe River downstream of the Thomson confluence. However, within the water column, the upper layer varies in thickness and prominence depending on distance from Lake Wellington. Often there is gradual transition rather than sharp boundary between the upper and lower layers. To maintain a consistent approach when interpreting the results of the hydrodynamic modelling, the in-channel bench was used to delineate the boundary between the upper and lower layers.

7ASM Full Paper

Lauchlan Arrowsmith et.al. – Salinity. Water Level, and Flow Considerations of the Lower Latrobe River

Figure 2. Definition of upper and lower layers within the water column Figure 3 shows the variation in salinity conditions for a range of flows throughout the estuary. Under prolonged low flow conditions, represented by inflows at or below 650 ML/day at the Swing Bridge for a period of 2 months, saline water in the form of a salt wedge was found to penetrate upstream into the Thomson River as far as 2.3 km from the confluence. This is a total of 17.3 km upstream of Lake Wellington.

In the upper part of the estuary, the halocline (interface between the fresh and saline waters) is quite clearly defined and occurs deep in the water column. In the mid estuary, upstream of Dowd Morass, the halocline is less sharply defined and can occur in the middle of the water column. In the lower estuary, the halocline is often not clearly defined due to mixing of the water column, and when it occurs it can be present high in the water column.

The rate of upstream movement of the salt wedge is dependent on both the river flows and the level and salinity of Lake Wellington. Observations of the movement of the salt wedge during the simulations indicate that upstream movement can also be affected by the river bed bathymetry. In locations where there is a distinct increase in bed level, such as at the confluence, the saline water movement slows and can be limited under typical low flow conditions.

After the salt wedge is flushed from the system by larger flows, the time it takes to move back into the estuary depends on the flow conditions in the river. If the flows remain above the various flushing thresholds it takes longer for the salt wedge to move upstream than under lower flow conditions.

Figure 3. Spatial Variability of Salinity under Variable Flow Conditions in the Latrobe estuary Three main flushing conditions (or “flushing criteria”) have been identified for the Latrobe estuary. In “Fully Flushed” conditions, the salt wedge is absent and both layers of water are fresh. In “Upper Flushed/Lower Part Flushed” conditions, the salt wedge is present in the lower layer. In “Not Flushed” conditions, brackish or saline water is mixed

7ASM Full Paper

Lauchlan Arrowsmith et.al. – Salinity. Water Level, and Flow Considerations of the Lower Latrobe River through the water column. For the different reporting locations in the estuary there are different flow conditions under which each of the three “flushing” criteria occurs, and these are summarized in Figure 4 for a range of flow scenarios; Natural inflows – no extraction, Inflows under the current level of development, Inflows under the full level of development (full uptake of extraction licenses), Natural inflows under climate change, Inflows under climate change under the current level of development, Inflows under climate change under the full level of development.

Figure 4. Summary of Salinity Flushing Conditions throughout the Latrobe Estuary

Environmental Water Recommendations

Overview Environmental water requirements for the estuary were determined based on the hydrological, hydrodynamic and salinity requirements for a range of environmental objectives identified in the study. For example, the natural riparian vegetation communities of the Latrobe estuary evolved in a predominantly freshwater environment and comprise of species that require fresh water to reproduce and thrive. Hence, fresh water is required in the root zone and during inundation events, particularly for riparian forest and woodland and paperbark woodland communities. Reed bed communities may be more tolerant of brackish conditions, although different populations of reed beds within the Gippsland Lakes vary in salt tolerance (Brizga et al, 2013).

The environmental water requirements were then translated into environmental flow recommendations in the same form as outlined in EEFAM (Lloyd et al. 2011). This translation of the physical, flow dependent habitat and ecological requirements into hydrological events and hydraulic processes is similar to the riverine FLOWS process. Thresholds must be identified which represent the minimum level at which the requirements are met. When the thresholds are known, the means of providing them through flow events can be specified.

This presented a difficult challenge in the case of the Latrobe estuary. The Latrobe estuary differs from ‘typical’ estuaries in that it forms part of a larger estuary (i.e. the Gippsland Lakes) and is situated immediately upstream of a lake (Lake Wellington), rather than the ocean. Water levels and salinity concentrations in Lake Wellington are more variable and less predictable than the ocean. This means that some ecologically important conditions in the Latrobe estuary (e.g. river level) do not readily translate to a single inflow rate. This makes the development of definitive flow recommendations more difficult than for typical estuaries.

7ASM Full Paper

Lauchlan Arrowsmith et.al. – Salinity. Water Level, and Flow Considerations of the Lower Latrobe River

For example, the relationship between water level and flow throughout the system is not clearly defined. In the upper sections of the study area, such as on the Thomson River, the river flow requirement to generate a given water level can vary by more than 10,000 ML/day. At the lower end of the system near Lake Wellington the water level in the estuary is largely independent of river flow, with the relationship primarily dependent on the water level of Lake Wellington.

Method In order to translate the environmental water requirements into environmental flow recommendations in the format similar to the Victorian FLOWS and EEFAM procedures, the hydrodynamic and salinity analyses were used to identify flow dependent processes that link ecological and geomorphic requirements. Where possible, the flow recommendations for environmental assets and objectives could then be specified in terms of flow components.

Flow components were specified flow events entering the estuary system. For the Latrobe estuary the following flow components were described as follows:

• Low flows – flows required to maintain freshwater in the upper water column (defined by the upper flushed/lower part flushed condition. The magnitude of the required low flow varies spatially throughout the estuary. • Freshes – flows required to completely flush saline water from the water column. As with low flows, the magnitude of the flushing flow requirements varies spatially throughout the estuary. • Bankfull and overbank events – again, bankfull and overbank flow event magnitudes vary spatially throughout the system. In addition, the water level related to these events can be driven by flow, lake water level of a combination of both. This means that for any given water level there is a wide range of corresponding flow conditions possible. In some locations in the lower estuary the water level is independent of river flow. • Asset specific flows – where flow dependence has been identified for a specific ecological or geomorphic asset that cannot be described in terms of low flows, freshes, bankfull or overbank flows these are specified as an individual flow component. The above terminology differs from FLOWS and EFAM due to the unique characteristics of the Latrobe estuary. In particular, the term “Low flows” is not defined as continuous base flows, but rather in terms of the flows required to maintain freshwater in the upper layer of the water column. The purpose of continuous base flows in non-estuarine reaches is to maintain aquatic habitat and hydraulic connectivity. Ambient freshwater habitat (in the upper layer of the water column) and hydraulic connectivity within the estuary are maintained by the non-continuous low flows provisions. Larger flushing flows rather than continuous base flows would be required to maintain freshwater habitat throughout the water column.

Conclusions Using detailed 1D and 3D hydrodynamic modeling techniques the environmental water requirements for the Latrobe estuary were determined based on the hydrological, hydrodynamic and salinity requirements to support the various environmental objectives. General and seasonal requirements in regard to flow, water level and salinity were defined. A key requirement for objectives relating to the natural freshwater ecosystems of the Latrobe estuary is the presence of fresh water (rather than brackish or saline water).

Environmental flow recommendations were developed relating to four main types of flow: A flushing flow regime sufficient to maintain fresh water in the upper layer of the water column, A flushing flow regime sufficient to maintain fresh water throughout the water column, Flow events to support particular environmental assets, such as riparian vegetation, perched freshwater wetlands or life history processes for fish, and Larger flows to maintain geomorphological processes.

Acknowledgments The authors would like to acknowledge the following members of the study team; Dr John Tilleard, Professor Paul Boon, Dr Adam Pope, Dr Nick O’Connor, Andrew McMahon, Darren Quin, as well as the West Gippsland Catchment Management Authority and the Department of Environment and Primary Industries.

7ASM Full Paper

Lauchlan Arrowsmith et.al. – Salinity. Water Level, and Flow Considerations of the Lower Latrobe River

References Alluvium (2009) Briefing Paper: Business Case for the long-term health of the Latrobe River. Traralgon, Victoria: Report P108086_R01 by Alluvium for the West Gippsland Catchment Management Authority Brizga SO, Lauchlan Arrowsmith C, Tilleard J, Boon P, McMahon A, O’Connor N, Pope A and Quin D (2013). Latrobe Estuary: Environmental Water Requirements Report. Water Technology Pty Ltd report to West Gippsland Catchment Management Authority Earth Tech (2005a). Assessment of Environmental Flow Requirements for the Latrobe River and Wetlands of the Lower Latrobe River – Site Paper. Report to West Gippsland Catchment Management Authority Ecos Consortium (2009). Understanding the environmental water requirements of the Gippsland Lakes System: Technical Report, Ecos Environmental Consulting, prepared for the West Gippsland Catchment Management Authority, Traralgon, Victoria. Gippsland Lakes Task Force, (2004). State of the Gippsland Lakes, Report prepared for the Gippsland Coastal Board Lloyd, LN., Anderson, BG., Cooling, M, Gippel, CJ, Pope, AJ and Sherwood JE (2011). Estuary environmental Flows Assessment Methodology: Final Specification Report. Lloyd Environmental Pty Ltd Report to Corangamite CMA, Colac, Victoria Longmore, A. R. (1990). Saltwater Intrusion into Lake Wellington, Gippsland Lakes, SRS 89/006, Marine Science Laboratories, Queenscliff, Victoria